Update to

A Spanish Repertoire for Black

by grandmaster Mihail Marin

15th February 2008

Quality Chess
www.qualitychessbooks.com
Theory moves on whether we like it or not. When a book is published on a variation the variation tends
to get more popular. This has happened with the closed Spanish over the last 12 months, in some cases
with a clear link to this book, including a game between a commending reviewer of the book and the
publisher, in other cases by the authors own practice.
For this reason we talked Mihail into doing an online update for his book. We hope that this is
received in the correct spirit and will be a pleasant surprise to fans of the book.
John Shaw and Jacob Aagaard, Quality Chess

Spanish Update

Update to Chapter 2

The Rubinstien System

The following can be read in connection with the
text on page 69 in the book.
Despite that ...e8 is entirely playable, as the
previous analysis has proven, abstract reasons
make me prefer ...d7. This developing move
involves a lesser commitment than the knight
retreat and allows recapturing with the queen in
some cases.
Op den Kelder-Marin, Banyoles 2007
15.axb5 axb5 16.b4 d7

1222222223
4 T M Tl+5
4+ WvVoOo5
4 + O M +5
4+oOpO + 5
4 P +p+ +5
4+ P +n+p5
4 +bN Pp+5
4R BqR K 5
7888888889

17.bxc5
Again, this is the most concrete way of trying to
refute Blacks strategy.
The merits of Blacks last move become
obvious after the risky 17.c4 when Black
can rapidly complete his development with
17...cxb4 18.cxb5 b7 19.b1 fc8

17...xc5
With the knights relatively far from the d6square, it makes little sense for Black to capture
with the pawn.
18.e3
White needs to spend a tempo on defending
this pawn. As indicated by Kortschnoj in his notes
to the game against Spassky, the straightforward
18.a3 xc3 19.e3 c7 20.xe5 leads
nowhere because of 20...xh3! It should be said
that Kortschnoj did not mention ...d7 as a
preparation to the continuation of Rubinsteins
plan but just as a possible way to delay ...b7
with one move.
After 18.e3, the threat a3 becomes serious.
18...c7
This is my recommendation from the previous
edition. Black removes the queen from the
exposed position, enabling the knights transfer
to c5 at the same time.
Previously
18...e8?!
had been played. It certainly looks like a logical
move. Black parries the threat by developing a
piece. However, in doing so he slightly neglects
the queen side situation, which could have led
to some problems.
19.c4!?
This move would have ensured White some
initiative.
In the game, White missed his chance to take
advantage of the exposure of the black queen
and focused on his kingside plan with 19.g4 g6
20.f1 b7 21.g3 a8 22.b1 ec8 23.e1
a5 24.f3 c4

Mihail Marin

1222222223
4t+t+ +l+5
4+ +vVo+o5
4 + O Mo+5
4+oWpO + 5
4 +m+p+p+5
4+ P +rNp5
4 +b+ P +5
4+rBqN K 5
7888888889

Black has managed to regroup in optimal way

and has a very enjoyable position, KuzminKrogius, Perm 1971.
19...b7
19...bxc4 leaves White with an active position
after 20.c3 b5 21.b1.
20.cxb5 xb5
20...ec8 would lose a tempo compared with a
similar line from the next comment, allowing
21.d3 since ...c5 is impossible.
21.b3
Black is not fully coordinated yet.
19.a3
The opening of the queen side by means of
19.c4 is not dangerous because of 19...b7!
(Development above all! 19...b4 is bad because
of 20.b3 when the weakness of the b4-pawn
prevents the knights activation.) 20.cxb5 fc8
for instance 21.a4 c5 22.b6 xb6 23.xd7
fxd7 24.ea3 d8!? ...g5.
19...e8
Now, this move becomes necessary.
20.b4 b7

1222222223
4 T +t+l+5
4+mWvVoOo5
4 + O M +5
4+o+pO + 5
4 B +p+ +5
4+ P Rn+p5
4 +bN Pp+5
4R +q+ K 5
7888888889

I played all these moves rather confidently,

because they were part of my analysis from the
first editions' main line. My opponent played
even quicker than me, though. I started fearing
that he had red the book and found a refutation
somewhere, although I trusted my analysis to
be correct. I had good reasons to take such a
possibility into account, since several opponents
from the previous rounds told me they had red at
least parts of my recent opening books.
Right after the game, when I asked my
opponent why he played so quickly he answered
rather vaguely, mentioning an older game he had
seen in the database. I was slightly disappointed:
no he did not read the book!
Now that I am writing these lines, the frustration
becomes even stronger. As can be seen below,
among the main actors of the modern phase from
the Rubinstein system's evolution we find the
author (kindly yours), one of the editors, without
whose permanent support the whole project
would have been impossible to accomplish (Jacob
Aagaard), and a famous reviewer, who had been
kind enough to write positively about the first
edition (Jonathan Rowson). The only missing
part is a reader...
21.b3
This is the new move compared to my analysis,
but during the game I could not remember
the variations very clearly. My initial line goes
21.a7 b6 22.a1 c5 eventually followed by
...a4.
21...f8
A necessary move. Black places the bishop
on a very stable square, offering to the e8-rook
freedom of action. I was worried that after
21...c5 22.a5 c8 23.xc5 xc5 24.d3 I
would not be able to regroup properly.
22.a7 b6
I did not feel prepared to fight for the a-file yet,
noticing that after 22...a8 23.a1 b8 24.xa8
xa8 25.e1 the occupation of the c5-square is
not possible without giving up the control of
the a-file. However, 25...h5 might offer Black
sufficient counterplay.
23.a1

Spanish Update

1222222223
4 T +tVl+5
4Rm+v+oOo5
4 W O M +5
4+o+pO + 5
4 B +p+ +5
4+nP Rn+p5
4 +b+ Pp+5
4Q + + K 5
7888888889

23...c5
Now, everything is ready for this long awaited
move.
24.e1
Avoiding the trap 24.xc5? dxc5 25.a6 b7
26.a7 a8!, winning material for Black.
24...a4
Black has neutralized Whites initiative,
achieving stability on the queenside. From the
openings point of view, he can be satisfied, but the
whole middlegame lies ahead. While writing the
first edition, I might have failed to emphasize how
difficult (for both sizes!) this phase of the game can
be, maybe because I was not completely aware of
it myself. In other words, in the long variations of
the Ruy Lopez it is not enough to equalize with
Black (or, similarly, get an advantage with White)
out of the opening. You also need to deal with the
strategic and tactical subtleties of these complex
positions properly.
I understood this truth with the occasion of my
recent win against Jakovenko, which can be found
in the 10.d5 Yates variation. Chess is enormously
complicated and the Ruy Lopez offers us a good
proof about it.
During the next phase of the game, Black had
tempting alternatives in several moments and even
now I am not sure whether my choices have always
been best. Maybe in this type of position there is no
such thing as the best move and it largely depends
on styles of play and taste. Since there is very little
practical material available in this variation, I have
taken myself the liberty to insert the whole game,
highlighting most of the critical moments and the
main alternatives in both sides play.

25.a5 h5
Blacks other knight goes to the edge of the
board, aiming to set up play on both wings. The
more cautious 25...a8, exchanging Whites most
active piece, was entirely possible.
26.bd2
I considered 26.d3 f4 27.f1 to be safer.
26...f4

1222222223
4 T +tVl+5
4+ +v+oOo5
4 W O + +5
4Ro+pO + 5
4mB +pM +5
4+ P +n+p5
4 +bN Pp+5
4Q + R K 5
7888888889

The pressure exerted by this knights is quite

annoying, but Black does not threaten anything
concrete yet.
27.c4
In the meantime, the other wing is under fire.
27...c5!
The knight had become unstable on a4. After
retreating to c5, it threatens to join his actions with
his colleague from the other wing for invading the
d3-square.
28.a3 ec8
Black brings another piece into play, refraining
from an early release of the tension. I saw that
Black could get a good position with the logical
and consequent 28...cd3 for instance 29.xd3
xd3 30.xd3 bxc4 31.xc4 (31.xc4 xb4
is a better version for Black because his queen
is more active.) 31...xb4 32.xb4 xb4=
However, I did not see an active plan in the final
position and decided not to part with my active
knights so easily.
29.b1
In order to maintain the balance even on the
queenside, White has to leave the e2-square
undefended.
29...d8

Mihail Marin

Once again, it was not easy to refrain from the

knight jump forward, but things are not entirely
clear after 29...e2 30.f1 d4 31.xd4 exd4.
For instance 32.a1 (The alternate way to attack
the d4-pawn would be 32.f3, but this allows
Black generate incontrollable complications with
32...bxc4 33.xc5 xb1 34.xb1 xb1 35.e2
dxc5, when Blacks pawns look scary.) 32...e7
33.xd4 f6 34.e3 a6!? and Blacks better
coordination offers him adequate compensation
for the pawn. From my choices in these two
critical moments we can define my general
approach in this phase: maintain the tension for
as long as possible, even though concrete action
seemed entirely viable, too.
30.h2 f6
I briefly considered 30...b7 31.a7 bxc4 but
did not like the idea of playing with a passive
knight on b7.
31.e3

1222222223
4 Tt+ Vl+5
4+ +v+oOo5
4 + O W +5
4RoMpO + 5
4 Bp+pM +5
4+ + Qn+p5
4 +bN PpK5
4+r+ + + 5
7888888889

31...e7!?
Again a quiet move, bearing in mind to transfer
the bishop to b6, which must be a familiar scenario
to the reader already. Instead of this somewhat
slow manoeuvre, Black had none less than two
worthy alternatives. I calculated 31...h5!? 32.h4
b7 33.a7 g6 34.g3 xd5 35.cxd5 xc2
36.b6 and considered it too risky. Maybe Black
has adequate counterplay after 36...g4 37.e1
xd2 38.xd2 xe4 but I was not sure at all.
31...b7 32.a7 bxc4 was possible, with the
possible continuation 33.xd6 xd6 34.axb7
xb7 35.xb7 c5 when Blacks dark squared
bishop gets into play faster than in the game.

32.a7
Immediately taking advantage of the temporary
vulnerability of the bishops.
32...b7
A necessary retreat already. However, the
passivity of the knight is temporary, because
Black has in mind the following plan: ...d8,
...g6-f8, ...c5 and ...b6. Quite slow, but the
vulnerability of Whites queen and queens rook
will eventually cause him some loss of time as well.
33.g1
Threatening with g3.
33...d8?!
This consequent move might not be best. Black
had an alternate way of activating his bishop with
33...g6 34.g3 g5 (I saw that 34...xd5 is
dangerous because of 35.cxd5 xc2 36.b6, when
36...d8? is insufficient in view of 37.xb7 xb6
38.xb8 c8 39.xb6- with decisive material
advantage for White.) 35.h4 h6 36.c3. For
some reason, I disliked this position, failing to
notice that Black can increase his pressure with
36...h5! (Threatening ... xd2.) 37.gf3 g4
when White would face problems keeping his
position together.
34.g3 g6 35.d3 f8 36.cxb5 c5

1222222223
4 TtV Ml+5
4R +v+oOo5
4 + O W +5
4+pMpO + 5
4 B +p+ +5
4+ +bQ Pp5
4 + N P K5
4+r+ + N 5
7888888889

White has won a pawn, but Black is just one

step away from crowning his previous play with
...b6.
37.xc5 xc5 38.a6
By this moment, my young opponent had
entered the phase of eternal time-trouble already
(with the 30 seconds increment), while I had
some spare minutes still.

Spanish Update

38...c3
Increasing the force of the threat ...xb5
and avoiding the trap 38...b6? creates an
unfavourable pin and can be answered with
39.c4! xc4 40.xb6
38...xb5!? was possible, though.
39.b6?!
39.b3 xb3 40.xb3 xb5 41.xb5 xb5=
looks safe for White because the bishop cannot
be activated easily.
39...xb6
I could not find adequate compensation for
the exchange after 39...xd3 40.xd3 xf2
41.h1 xb6 42.f1 and had no time left to
notice the simple 39...c8!?
40.axb6 xd3 41.e2 xb6 42.xb6 c3

1222222223
4 + + Ml+5
4+ +v+oOo5
4 R O W +5
4+ +pO + 5
4 + +p+ +5
4+ T + Pp5
4 + NqP K5
4+ + + N 5
7888888889

After a long tense phase, play has calmed down

somewhat. Black's position is more compact and
his king safe. The e4-pawn is more vulnerable
than the d6-pawn, because f3 would weaken the
king even more.
43.g2
Preparing c4, which was impossible now
because of 43.c4 xc4 44.xc4 xf2 45.h1
xg3 with two pawns for the exchange and
threats against the enemy king.
43...h6
Slightly careless. I had practically no time left
either and played a generally useful move, failing
to notice Whites reply. A safer way to reach a
similar position as in the game was 43...d8!
44.a6 (44.xd6? loses to 44...xh3 while after;
44.b4 Black can play 44...h6 already.) 44...c2
45.gf3 a2!. Releasing the pressure against the

d6-pawn. 46.xa2 xb6

44.c4 d8!
Fortunately, this still works out well.
45.b4
45.xd6 loses the exchange to 45...c7
45...c7 46.e3 h7
The wandering knight returns to its previous
location, in order to put the e4-pawn under
pressure.
47.c4
Hoping to ease his defence by exchanges.
47...xc4 48.xc4 b6 49.e2 f6 50.c3
d4

1222222223
4 + + +l+5
4+ +v+oO 5
4 + O M O5
4+ +pO + 5
4 +qWp+ +5
4+ N N Pp5
4 + + Pk+5
4+ + + + 5
7888888889

Black has completed his regrouping, and his

pieces dominate the position. Although it is
early to speaker about a concrete advantage yet,
White's defence in time trouble is not easy.
51.c7 h7
A prophylactic move, passing to the opponent
the responsibility of concrete action.
52.c2?

1222222223
4 + + + +5
4+ Qv+oOl5
4 + O M O5
4+ +pO + 5
4 + Wp+ +5
4+ N + Pp5
4 +n+ Pk+5
4+ + + + 5
7888888889

Mihail Marin

An understandable blunder. The queen's

presence in the centre was irritating...
52...xh3!
A decisive little combination.
53.xh3 xf2 54.b4?!
Saving the knight, but allowing a forced mate.
54...f1 10
We have enough elements to conclude that, against
both ...e8 and ...d7, White gets very little by
opening the queenside immediately. Therefore,
we should investigate the plan consisting of the
standard knight transfer to the kingside.
Berescu-Marin
Romanian Championship, Predeal 2006
15.axb5 axb5 16.b4 d7
This position can be reached via the move order
16...e8 17.f1 d7, too. In this latter variation,
17...g6 is slightly premature. In order to maintain
the re-capture on c5 with the pawn viable, Black
should delay the moment of weakening of the f6and h6-squares for as long as he has other useful
moves available. Now, or on the next move,
White could switch back to the plan 18.bxc5
dxc5 19.c4. A later comment will reveal the fact
that against Blacks correct move order this plan
is harmless.
17.f1

1222222223
4 T M Tl+5
4+ WvVoOo5
4 + O M +5
4+oOpO + 5
4 P +p+ +5
4+ P +n+p5
4 +b+ Pp+5
4R BqRnK 5
7888888889

17...e8
Black continues regrouping in the spirit of the
Rubinstein system.
18.e3 f6

Intending ...f7 and only later ...g6 and

...g7.
19.3h2
Now, the opening of the queen side by means
of 19.bxc5 dxc5 20.c4 would be less effective
because of the simple 20...bxc4 when the time
needed by White to win the pawn back could
be used by Black to transfer one of his knights
to d4. Here is a possible continuation: 21.a4.
This move carries out a strategically favourable
exchange but loses even more time. 21...xa4
22.xa4 d6. Please notice that this move is
enabled by the fact that the e5-pawn is safely
defended by his colleague. 23.c2 b4= and
Black has little to complain about. After the
careless 24.3d2? Blacks activity would become
threatening starting with 24...c3!
19...f7 20.g3 g6 21.d2 g7
Both sides have completed the first phase of
piece mobilization. The exchange on c5 does not
offer White anything because after ...dxc5 Black
would get the excellent d6-square for his minor
pieces.
22.f4
Otherwise, Black could start active kingside
operations himself.
22...exf4 23.xf4

1222222223
4 T + Tl+5
4+ WvVmMo5
4 + O Oo+5
4+oOp+ + 5
4 P +pB +5
4+ P + Np5
4 +bQ +pN5
4R + R K 5
7888888889

23...a8
Black prevents his opponent from taking
over the control of the a-file, but makes a slight
structural concession. 23...e5 would have
been more ambitious from strategic point of
view, although chances would have remained
roughly equal after 24.a2 a8 25.ea1 xa2

Spanish Update

26.xa2 b8= The strong centralized knight

and the flexibility of Blacks queenside structure
compensate for Whites control of the only (yet!)
open file.
24.bxc5 xc5 25.e3 c7 26.d4

1222222223
4t+ + Tl+5
4+ WvVmMo5
4 + O Oo+5
4+o+p+ + 5
4 + Bp+ +5
4+ P + Np5
4 +bQ +pN5
4R + R K 5
7888888889

White has obtained this wonderful square

for his minor pieces, but Black will solve all his
problems by means of simplifications along the
a-file.
26...e5 27.hf1 b7 28.b3 xa1 29.xa1
a8 30.xa8 and a draw was agreed.
The game was played shortly after I had
delivered the final form of the book to my editors,
but months before its publishing. My opponent,
who had won the Romanian championship
twice over the last years and is regarded as a
good theoretician, expressed his disappointment
regarding the ease with which Black managed to
equalize, using an officially doomed variation.
When I mentioned the fact that there will be a
70pages chapter dedicated to the whole system,
he relaxed and stopped feeling uncomfortable.
At the same time, I felt quite happy because
I managed to defend Blacks point of view in
a practical game. Analyzing long variations is
good, but over-the-board testing should be an
important element, too.
Whites failure to obtain an advantage was
mainly caused by Blacks potential threat of
taking over the initiative on the kingside with
...f5. This forced him to open the position with
f4 at an earlier stage than he might have wished
to. Therefore, after Blacks
17...e8,

the ambitious
18.g4
deserves being investigated.

1222222223
4 T MmTl+5
4+ WvVoOo5
4 + O + +5
4+oOpO + 5
4 P +p+p+5
4+ P +n+p5
4 +b+ P +5
4R BqRnK 5
7888888889

On the dark side of this move I would mention

that it implies a serious kingside commitment
from White, which basically excludes the
possibility of successfully opening the other wing
with bxc5, allowing Black play
18...g6
with all the comfort. For a while, play continues
in accordance with the familiar patterns.
19.g3 f6 20.e3 f7 21.d2!
A natural move, connecting rooks and preparing
for active operations on both wings. In the only
relevant game played prior to the first edition,
21.a3 was tried. After 21...g7 22.a1 c8!
A subtle move, preparing active operations on
both wings. Black increases the probability of the
thematic break ...f5, while enabling the activation
of the dark-squared bishop via d8-b6. 23.a7
d8! Black had a good position in TukmakovKan, Tashkent 1974. After the transfer of the
bishop to the queenside, White has little chances
for success on this territory. The game continued
24.a6 (The tactical justification of the last move
is that 24.bxc5 dxc5 25.xc5 xc5 26.xd7
leaves the rook trapped to 26...b6 27.f1 c8.
In the first edition I gave 26...b4, which is also
good and offers Black excellent compensation for
the pawn, in view of his control over the dark
squares. Safer would have been 24.c1 although
after 24...b6 25.a6 c7 Black has little to fear.)
24...f5! (A well timed pawn break, after which
White will not be able to justify his multiple

Mihail Marin

commitments on different areas (g4, b4, the

occupation of the a-file). White probably hoped
for 24...cxb4?! 25.cxb4 xc2?! which loses the
queen after 26.c1 b3 27.d2 xb4 28.b1)
25.exf5 gxf5 26.gxf5? (This moves weakens the
d5-pawn, which will be quite relevant as we shall
see. White should have embarked the dangerous
complications starting with 26.bxc5 f4) 26...
cxb4! 27.a7 (The difference is that after 27.cxb4
xc2 28.c1 b3 the queen will escape through
d5.) 27...bxc3 Black had material and positional
advantage and won soon. Curiously, this game
escaped the theoreticians attention even though
it was played at a time when the variation was
officially alive.
21...g7 22.a2
White plans to increase his pressure by doubling
rooks along the a-file. For the time being, Black
cannot initiate a kingside counterplay with
...f5 and his knight has not the e5-square at his
disposal. He has to look for a way to consolidate
his position, maintaining the hope that the
weakness induced by the courageous advance of
Whites g-pawn will tell at a later stage. We shall
investigate two possible continuations.

1222222223
4 T + Tl+5
4+ WvVmMo5
4 + O Oo+5
4+oOpO + 5
4 P +p+p+5
4+ P BnNp5
4r+bQ P +5
4+ + R K 5
7888888889

22...c8
The same plan as in the game TulmakovKan, although the tactical nuances are slightly
more complicated here because of the different
placement of the white queen.
I believe that it is best to play this move
immediately. Alternatively, Black can trade a pair
of rooks along the a-file with 22...a8 23.ea1
xa2 (In the game Aagaard-Rowson, British

Championships 2007, Black continued the fight

for the a-file with 23...b7?! White answered
with the strong move 24.b3!, setting up indirect
pressure along the a2-g8 diagonal. Blocking the
position with 24...c4 would open the diagonal of
the e3-bishop, allowing the intermediate 25.a7!
The prophylactic 24...h8 looks safer, although
it is not easy to suggest a further plan for Black.
The game went 24...xa2?! 25.xa2 c8? 26.bxc5
dxc5 27.d6! with strong initiative for White.)
24.xa2 c8! 25.a7 d8 26.bxc5. Otherwise,
Black would be very much OK after ...b6. 26...
dxc5 27.xc5 xc5 28.xd7

1222222223
4 + V Tl+5
4+ +r+mMo5
4 + + Oo+5
4+oWpO + 5
4 + +p+p+5
4+ P +nNp5
4 +bQ P +5
4+ + + K 5
7888888889

White has won a pawn, but Black has

good chances to stabilize the position and
setup domination on dark squares. Now, the
weaknesses induced by g4 are obvious. 28...e8
(This looks to be the most solid continuation,
keeping the d6-square under firm control and
preparing an optimal regrouping of the knights.
28...a5 is possible, but can lead to unnecessary
complications after 29.e2 e8 30.e3!? It is
useful to exchange the active black queen, even
at the cost of weakening the own structure.
30...xe3 31.fxe3 ed6. Blacks position looks
very nice in view of the threats ...b6 followed by
...c4, but White has not exhausted his resources
yet. 32.g5!? Played with the hope for 32...g7?
33.ed4!! exd4 34.gxf6 xf6 35.e5!. A nice
tactical resources, but Black can do better. 32...
fxg5 33.xe5 xe5 34.xd6 f3 35.h1 e1
with a very unclear position, where Whites
pieces lack coordination and are in fact hanging.)
29.e3 xe3 30.fxe3 ed6. Now, the threat

10

Spanish Update

...a5 is real and 31.g5?! fxg5 32.xe5? does not

work anymore because of 32...xe5 33.xd6
c7 34.e6 f3 winning material.
23.a7
White tries to prevent Blacks plan by tactical
means. In case of the more neutral 23.ea1 Black
achieves a safe position with 23...d8=
23...d8 24.bxc5
The logical consequence of the previous move.
24...dxc5 25.xc5 xc5 26.xd7
The b8-rook is useful by over-defending the
b5-pawn and depriving the enemy rook of the
b7-square. In some cases, it enables the bishops
transfer to b6. These are sufficient reasons to justify
refraining from exchanging rooks along the a-file
before initiating the thematic regroupment.
Black has several ways to obtain very good
play.
26...e8
26...a5 is also strong, when 27.e2? loses
material to 27...b6 28.g3 c8 29.e7 d8
27.d3
27.e3? is impossible now because of 27...c8!
28.a7 b6 with a deadly fork.
27...ed6

1222222223
4 T V Tl+5
4+ +r+m+o5
4 + M Oo+5
4+oWpO + 5
4 + +p+p+5
4+ Pb+nNp5
4 + Q P +5
4+ + R K 5
7888888889
If anything, I would prefer Black here. The d7rook is vulnerable, while Black's position is very
stable.
We have now reached the end of an incredibly
long journey. I hope that I have managed to
prove that the system of development designed
by Rubinstein one century ago is perfectly
playable and that the oblivion into which it has
fallen for more than 3 decades is due only to the

Theory
So footnote 155 is not perfect. More accurate
would be: 16.f1 e8 17.f1 d7 transposes
to 15d7. 17...g6?! 18.bxc5!? dxc5 19.c4.
Line 27 in the book could be replaced with the
three following lines: A, B and C.
Line A
15d7 16.axb5 axb5 17.bxc5
17.c4 cxb4 18.cxb5 b7 19.b1 fc8
17...xc5 18.e3
Threatening a3. 18.a3 xc3 19.e3 c7
20.xe5 xh3! Kortschnoj.
18...c7
18...e8?! 19.c4!? (19.g4 g6 20.f1 b7
21.g3 a8 22.b1 ec8 23.e1 a5 24.f3
c4 Kuzmin-Krogius, Perm 1971.) 19...b7
(19...bxc4 20.c3 b5 21.b1) 20.cxb5 xb5

(20...ec8 21.d3) 21.b3

19.a3
19.c4 b7! (19...b4 20.b3.) 20.cxb5 fc8
21.a4 c5 22.b6 xb6 23.xd7 fxd7
24.ea3 d8!? ...g5.
19...e8 20.b4 b7 21.b3
21.a7 b6 22.a1 c5 eventually followed
by ...a4.
21...f8
21...c5 22.a5 c8 23.xc5 xc5 24.d3.
22.a7 b6
22...a8 23.a1 b8 24.xa8 xa8 25.e1
(a)25...h5.
23.a1 c5 24.e1
24.xc5? dxc5 25.a6 b7 26.a7 a8!
24...a4=
Black has neutralized Whites initiative,
achieving stability on the queenside, Op den
Kelder-Marin, Banyoles 2007.

Mihail Marin

Line B
17.f1 e8 18.e3 f6
...f7, ...g6, ...g7.
19.3h2
19.bxc5 dxc5 20.c4 bxc4 21.a4 21...xa4
22.xa4 d6 23.c2 b4= 24.3d2? 24...c3!
19...f7 20.g3 g6 21.d2 g7
...f5
22.f4 exf4 23.xf4 a8
23...e5 24.a2 a8 25.ea1 xa2 26.xa2
b8=
24.bxc5 xc5 25.e3 c7 26.d4 e5
27.hf1 b7 28.b3 xa1 29.xa1 a8
30.xa8 draw,
Berescu-Marin,
Romanian
championship,
Predeal 2006.
Line C
18.g4
Controlling f5, but weakening f4.
18...g6 19.g3 f6 20.e3 f7 21.d2!
21.a3 g7 22.a1 c8! (...f5, d8-b6.)
23.a7 d8!= 24.a6 (24.bxc5 dxc5 25.xc5
xc5 26.xd7 b6 27.f1 c8 or 26...b4.
24.c1 b6 25.a6 c7=) 24...f5! (24...cxb4?!

11

25.cxb4 xc2?! 26.c1 b3 27.d2 xb4

28.b1) 25.exf5 gxf5 26.gxf5? (26.bxc5 f4)
26...cxb4! 27.a7 (27.cxb4 xc2 28.c1 b3
...d5) 27...bxc3- Tukmakov-Kan, Tashkent
1974.
21...g7 22.a2 c8
22...a8 23.ea1 xa2 (23...b7?! 24.b3!
xa2?! 25.xa2 c8? 26.bxc5 dxc5 27.d6!
Aagaard-Rowson, British Championships 2007.
24...c4 25.a7! 24...h8!?) 24.xa2 c8!
25.a7 d8 (...b6) 26.bxc5 dxc5 27.xc5
xc5 28.xd7 e8 (28...a5 29.e2 e8
30.e3!? xe3 31.fxe3 ed6. ...b6, ...c4.
32.g5!? g7? 33.ed4!! exd4 34.gxf6 xf6
35.e5!. 32...fxg5! 33.xe5 xe5 34.xd6 f3
35.h1 e1) 29.e3 xe3 30.fxe3 ed6.
...a5. 31.g5?! fxg5 32.xe5? xe5 33.xd6
c7 34.e6 f3 -.
23.a7
23.ea1 d8=
23...d8 24.bxc5 dxc5 25.xc5 xc5 26.xd7
e8
26...a5!? 27.e2? b6 28.g3 c8 29.e7
d8
27.d3
27.e3? c8! 28.a7 b6.
27...ed6=

12

Spanish Update

12.d5
In the Chigorin variation, instead of the natural
developing move 12.bd2, the immediate

1222222223
4t+v+ Tl+5
4+ W VoOo5
4o+ O M +5
4MoO O + 5
4 + Pp+ +5
4+ P +n+p5
4pPb+ Pp+5
4RnBqR K 5
7888888889

12.d5
is also possible.
This move has become popular in recent
years, especially among players who like to avoid
the systems based on an early exchange on d4.
From our point of view it does not induce any
significant differences from the main lines.
12...d7
This is the most flexible move. After the
hurried 12...c4 White could play 13.b4! b7 (if
13...cxb3 14.axb3 Whites undeveloped knight
proves useful, because the c3-pawn is defended)
14.a4 when, compared to the Rubinstein system,
Whites queens knight could prove useful in
attacking the b5-pawn.
13.b3
White intends to attack the b5-pawn with a4
and a3.
The immediate 13.a4 would weaken the
b3-square, allowing 13...c4! in favourable
circumstances.
The most consistent continuation is 13.bd2,
but this would transpose to the second line from
the Petrosian system tables.
13...ab8!?

Black aims to discourage the planned a4.

However, other moves such as 13...b7 and
13...fe8 are entirely playable, too.
14.e3
14.a4?! is not recommendable because of 14...
bxa4 with pressure against the b3-square.
14...g6
A move order that limits Whites options more
is 14...e8 15.bd2 g6 16.h6 g7.
15.bd2
White could have tried 15.h6!?, although then
Black would play in the spirit of the Petrosian
system with 15...fe8 16.bd2 f8=.
15...h5 16.h6 g7 17.f1 b7
The picture is very familiar for adherents of the
Rubinstein system.
18.g4 d8 19.g3 f6 20.h2 f7 21.e3
h8
21...a5!?
22.d2
22.f4 f5!? h4
22...c8
22...a5!?

1222222223
4 Tw+ T L5
4+ +vVmMo5
4o+ O Oo+5
4+oOpO + 5
4 + +p+p+5
4+pP B Np5
4p+bQ P N5
4R + R K 5
7888888889

23.f4! exf4
23...f5 24.exf5 gxf5 25.g5!
24.xf4 e5=
Maze Marin, Moscow Aeroflot 2005.

Mihail Marin

13

Update to Chapter 8

The Worrall Attack

Finally, I will mention a slightly tricky move
order, advocated by Tiviakov.
After 5.e2 b5 6.b3 e7, White can play
7.c3:

1222222223
4t+vWl+ T5
4+ OoVoOo5
4o+m+ M +5
4+o+ O + 5
4 + +p+ +5
4+bP +n+ 5
4pP PqPpP5
4RnB K +r5
7888888889

In the first edition, I have failed to notice the

significant differences induced by this move
and have not mentioned at all, assuming that it
would simply transpose elsewhere! Fortunately, I
got my punishment just in time to become aware
of the necessity to examine this variation for
the present edition. I suspect that my opponent
(Vladimir Baklan, an extremely efficient player
with the white pieces) or one of his team mates
had noticed my omission. Otherwise I cannot
explain why a strong player with a stable opening
repertoire chose precisely the Worrall Attack for
the first time in his life!
Be it as it may, during the game I was not aware
of the fact that there was no mention of 7.c3 in
the book and played 7...d6 thinking that after
8.d4 g4 we would transpose to the line 7.d4 d6
8.c3 g4.
Only when he answered 8.a4, did I remember an
important detail, which served me as a guideline

when writing the chapter for the first time, but

which I had failed to mention explicitly: Black
should not weaken the c6-square before castling,
in order to be able to meet a4 with ...b4 without
fearing xf7 followed by c4 or simply c4
with a double attack.

1222222223
4t+vWl+ T5
4+ O VoOo5
4o+mO M +5
4+o+ O + 5
4p+ +p+ +5
4+bP +n+ 5
4 P PqPpP5
4RnB K +r5
7888888889

Caught by surprise, I did not want to make

concessions such as 8...d7 or 8...b8, which
would offer White a more favourable form of the
normal line after castling followed by d4. Indeed,
the development of the bishop to d7 would allow
White spare the move h3, while abandoning the
a-file would disable (after the exchange on b5)
the typical manoeuvre ...a5. Instead, I tried
to repair my mistake from a position of force
with 8...g4, but, not being prepared for such a
course of events, went down rather painfully. My
highest praise and sincere thanks for Vladimir
and his accurate play in the decisive part of this
rather short game. Without his help, the book
you are reading would have been incomplete.
Further analysis proved that Blacks pawn
sacrifice is not entirely unsound, but my feeling
is that the variation does not fit in our main

14

Spanish Update

repertoire too well (Remember? We are playing the

Chigorin Variation, not the Marshall Attack!)
Immediately after resigning, I put all the
blame on the careless advance of the d-pawn
and decided that 7...00 would have avoided
troubles. This simplistic attitude certainly served
me well for avoiding a sleepless night and being
fit for the last and decisive round. Later, however,
I understood that things are not that simple and
that additional work has to be done. After 8.d4
d6 White is not really forced to transpose to the
previously investigated lines by castling, but can
consider consolidating his centre with either
9.bd2 or 9.h3.

1222222223
4t+vW Tl+5
4+ O VoOo5
4o+mO M +5
4+o+ O + 5
4 + Pp+ +5
4+bP +n+ 5
4pP +qPpP5
4RnB K +r5
7888888889

Personally, I am not too worried by 9.bd2 After

9...exd4 (This slight deviation from the Yatessystem plan is better than the immediate 9...g4,
which can be answered with 10.h3) 10.cxd4 g4
the early (premature!?) development of the queen
makes itself felt. The threat 11...xd4 forces the
white Lady move again, on a rather unnatural
position with 11.e3
Although there is no immediate danger for
White, he will face problems completing his
queenside development, allowing Black generate
adequate counterplay against the mighty central
pawns. In practice, Black has tried 11...d5 12.e5
e4 mainly, but I fear that closing the centre
would offer White the possibility to get his forces
coordinated. I prefer the plan suggested against
the e3 line in the Yates system, namely 11...a5
12.c2 c5

Whites more consistent continuation is 9.h3

1222222223
4t+vW Tl+5
4+ O VoOo5
4o+mO M +5
4+o+ O + 5
4 + Pp+ +5
4+bP +n+p5
4pP +qPp+5
4RnB K +r5
7888888889

The main difference compared to the line with

an early white castle is that White has occupied
the centre and secured his domination in this area
before Black carried out the Chigorin manoeuvre
...a5 and ...c5. This will make it more difficult
for Black to get active counterplay and certainly
poses us concrete problems maintaining the
coherence of our entire repertoire against the Ruy
Lopez. Later, White will place his rook on d1,
creating an unpleasant pin along the d-file and
eventually play a4, when the queens presence on
e2 would prove useful.
In practice, Black has been relatively successful
with 9...b7 10.00 e8 eventually followed
by ...d8 and ...b8-d7. While this hardly has
anything in common with the Chigorin system,
I am also slightly suspicious about the objective
merits of such a way of regrouping. With the rook
captive on f8 and the d8-bishop obstructing the
communication between Blacks major pieces, it
certainly looks like a worse version of the Breyer
system. Finally, It does not look like a serious
reason why not to play the Worral...
When searching for the optimal plan for Black,
I have focused on two main goals: to make the
queen feel uncomfortable on e2 and create some
similarity with the Rubinstein or the Petrosian
system. In order to enable the knight jump ...a5
followed by ...c5, Black obviously needs to overprotect the e5-pawn.
The most natural way to do it is 9...e8 10.00
10.g5 is not dangerous. After the calm
10...f8 the early departure of the queen from

Mihail Marin

d1 makes itself felt, by leaving the d4-pawn

insufficiently defended. In order to avoid an
unfavourable release of the tension in the centre,
White has nothing better than return with the
knight to f3, when Black can insist with ...e8.
11.e3 is dangerous because of 11...d5! when
Whites minor pieces are slightly hanging.
Besides, Black can try the more adventurous
10...d5, implying material sacrifices for the sake
of preventing the enemy king from castling.
10...f8

1222222223
4t+vWtVl+5
4+ O +oOo5
4o+mO M +5
4+o+ O + 5
4 + Pp+ +5
4+bP +n+p5
4pP +qPp+5
4RnB +rK 5
7888888889

By regrouping in the spirit of the Zaitsev and

the Smyslov systems, Black has immediately put
the e4-pawn under pressure.
Curiously, this plan has been played very rarely.
When confronted with this variation, practically
all the top players stuck to systems of development
that made part of their main repertoire against
5.00 followed by 6.e1 and which do not
necessarily work out equally well here. It is
possible that they were caught by surprize by the
move order based on 6.c3 and just looked for an
over-the-board emergency exit.
If this is the truth, I can understand them
perfectly well. Even in the peace of my working
place, I needed quite some time to find a solution.
I even tried to imagine the situation in which,
preparing for an opponent who frequently
plays 5.e2 followed by 6.c3, I had spotted my
omission in time. Would I have been able to find
my way out under the pressure of time? Probably
not; I might have simply switched to the Pirc,
which would have been a good practical decision
as a tournament player, but basically a desertion

15

as an author. See the similar situation described

in the Evans chapter.
Anyway, the lack of practical examples in this
line gives this part of the chapter a somewhat
vague character. However, I believe that any
player mastering the first chapters of the book
and my recommendations from the current
position will have no problems defending Blacks
point of view.
Let us consider Whites main continuations.
11.d5 is hardly a solution now. After 11...a5
12.b3 c6, Black gets a very favourable form of
the Yates system with 10.d5. Instead of losing
two tempi with the queens bishop (...g4 and
...c8) he has played two useful moves (...f8
and ...e8), which make part of his main plan
anyway! The fact that e2 had been played
instead of e1 hardly favours White.
11.g5 continues to be inoffensive. Black
answers 11...e7, maintaining the pressure
against Whites centre and leaving the knight
misplaced on g5.
11.g5 has been played, but Whites initiative
has only temporary character after 11...h6
12.xf6 xf6 13.d5 d7.
The most natural continuation is 11.bd2
when nothing can stop Black to play 11...a5
12.c2 c5:

1222222223
4t+vWtVl+5
4+ + +oOo5
4o+ O M +5
4MoO O + 5
4 + Pp+ +5
4+ P +n+p5
4pPbNqPp+5
4R B +rK 5
7888888889

Here it is, an almost typical Chigorin position!

Having played ...e8 and ...f8 prevents Black
from employing Rubinsteins regrouping already.
However, these moves are useful from the point
of view of the Petrosian system. Besides, the
presence of the white queen on e2 and Her

16

Spanish Update

Majestys opposition with the black rook will force

White release the tension in the centre earlier
than generally desirable. Just take into account
that Black has not made any committal move yet
(such as ...c6) and neither d5 nor dxc5 offer
White any concrete advantage.
In the Chigorin variation, the black queens
development on c7 was a concrete necessity
at a concrete moment, but not really a great
achievement for Black. Therefore, maintaining
other options for this mighty piece in the
diagrammed position offers Black greater
flexibility.
Finally, White has 11.a4:

1222222223
4t+vWtVl+5
4+ O +oOo5
4o+mO M +5
4+o+ O + 5
4p+ Pp+ +5
4+bP +n+p5
4 P +qPp+5
4RnB +rK 5
7888888889

This move caused Black some trouble in the

few games where it was played. White indirectly
defends the e4-pawn, because 11.exd4?! can be
strongly answered with the intermediate move
12.axb5!

The most flexible continuation is 11...d7,

indirectly defending the b5-pawn and maintaining
the same threats and ideas as before.
After 12.bd2 a5 13.c2 c5 14.d1 c7
we can notice that the advance of the a-pawn is
anything but one-sided.

1222222223
4t+ +tVl+5
4+ Wv+oOo5
4o+ O M +5
4MoO O + 5
4p+ Pp+ +5
4+ P +n+p5
4 PbNqPp+5
4R Br+ K 5
7888888889

15.f1 can be met with 15...exd4 16.cxd4

bxa4!, while the opening of the a-file with
15.axb5 axb5 before playing 16.f1 allows the
simplifications 16...b3! 17.xa8 xc1 18.xc1
xa8. In order to avoid repeating what I have
said so many times already, I will just offer you a
quiz: which is the drawback in Whites position
that enabled this favourable for Black tactical
operation?
I will preserve the conclusion from the first
edition (below), adding only that 6.c3 should not
allowed to be an over the board surprise, as I had
the occasion to learn on my own...

Theory
Line A
7.c3 d6 8.a4 g4 9.axb5
9.h3 a5 10.xf7?! (10.c2 xf3 11.xf3
leaves White a tempo down if compared with the
lines from chapter 10 - the a4-system.) 10...xf7
11.hxg4 b3 12.axb5 xa1 13.g5 Whites attack
is less dangerous than it looks at first sight. he is a
whole rook down and is rather poorly developed.
13...g4!? 14.c4 (14.h4 xg5 15.xg4
xh4 16.xh4 b3) 14...f8! 15.bxa6

(15.g6 e8! 16.xh7 g8) 15...c2

(15...xg5 16.a4) 16.e2 (16.f1 c8)
16...d5 17.exd5 c8
9...axb5 10.xa8 xa8 11.xb5 00 12.e2!
12.00 would transpose to a famous game that
I had known for more than 30 years. 12...a7!
(12...xe4? 13.d5 Keres; 12...b8 13.a4
Keres; 12...a5 13.c2 xe4 14.xe5! BookCHOD Alexander, Margate 1938.) 13.e2
(13.a5 xe4 14.xa7 xf3 15.gxf3 g6
16.h1 xb1 17.xc7 d3 Blacks initiative

Mihail Marin

should reach for at least perpetual check.)

13...xe4 14.xe4 xe4 15.d4 (15.d1 c5
Keres) 15...xf3 16.gxf3 g5 17.g2 (17.xg5
xg5 Keres) 17...b8 18.c4 exd4 19.cxd4
e6 20.d5 (20.xe6 fxe6 Keres; 20.e1 f6
21.xe6 fxe6 22.xe6 xd4 Keres) 20...c5
Fine-Keres, AVRO 1938.
12...b8
12...h5?! 13.d3 a1?! 14.00! (14.c2 f4
15.xf4 exf4 16.b4 e5) 14...b8 15.d5
d8 16.a3 Baklan-Marin, Romanian team
championship, 2007.
13.c2 h5 14.00
Paradoxically, this brave move seems to be
Whites best chance for an advantage.
14.d3 f4 15.xf4 exf4 16.b4 a2 followed by
...e5 or immediately 16...e5, with initiative
for the pawn.
14.g3 This move prevents the knight jump,
but leaves White with problems getting rid of the
unpleasant pin and get castled at the same time.
14...a2 15.h3 (15.00 a5 will more or less
force White give up his light-squared bishop for
the knight after 16.d3 b3, when the kingside
weaknesses would become more relevant. Or
16.d4 c4 and the queenside is in danger.)
15...d7 16.f1 (16.d4 exd4 17.cxd4 b4)
16...a5
14...f4
14...g5!? 15.g3 h6 is an interesting
alternative, leaving White with problems
completing his development.
15.e3 a5 16.d3 f5!? 17.b4
17.bd2 g5! 18.xg5? e2 19.h1 f4.
17...fxe4 18.xe4 xe4 19.dxe4 e2 20.h1
xc1 21.xc1 xf3 22.gxf3
22.bxa5 e2 23.d2 g5 24.b1 a8 25.b3
c5
22...c6
White has managed to keep his extra-pawn,
but his structure is far from perfect. Besides, the
presence of opposite coloured bishops and the
fact that Black will occupy the only open file
increases the probability of a draw dramatically.
For instance
23.a3 f8 24.c4 a8 25.b3 g5 26.e3
e7
and Black should not lose.

17

Line B
7...00 8.d4 d6 9.bd2 exd4 10.cxd4
10.xd4 xd4 11.cxd4 c5 12.dxc5 dxc5
13.00 c4
10...g4 11.e3
11.d3 b4 12.c3 c5 13.00 xf3
11...a5
11...d5 12.e5 e4 13.00 (13.xe4 dxe4
14.xe4 xf3 15.gxf3 xd4 16.d5 c2)
13...f5 14.c2 (Later, Tiviakov switched to
14.d1 d7 15.f1 a5 16.c2 c4 17.e2 f6
18.e3. Now, in the game Tiviakov-Ibragimov,
Elista 1997, Black should have played 18...xe3!?
19.xe3 g4 20.d3 e6 with an entirely
acceptable position.) 14...xd2 15.xd2 e4
16.d1 d7 (16...b4 17.b1 c5!? Beliavsky)
17.e2 f5 (17...g4! 18.h3 g6 Beliavsky)
18.e1 Tiviakov-Beliavsky, Cacak 1996.
18...xc2 (18...d8? 19.b3! f3 19...f4?
20.xe4!-) 19.xc2 d8 20.d3 e6 21.f4
ac8 ...c5, ...xf4, g5.
12.c2
12.00 should also be answered with 12...
c5, when the bishop would have to retreat later
anyway.
12...c5 13.00
13.d5 e8 14.00 f8 (...xd5.) 15.f4 g6
16.h3 xf3 17.xf3 This queen has moved quite
a lot. It would be illogical to claim an advantage
for White. 17...g7 18.b1 c8= 19.b3 c4!
20.b4? c3!
13.dxc5 dxc5 14.00 (14.e5? d5 15.e4
f5 16.exf6 xf6 or 15.xh7!? h8! leave
Whites pieces hanging.) 14...h5 (Anticipating
e5 by enabling ...g6 as an answer to a later
e4.) 15.b3 c6 16.b2 b4 17.b1 e8=
White has an advantage of space in the centre,
but his development is far from harmonious.
13...c8
If given a choice, I would prefer maintaining
the tension for as long as possible.
However, 13...cxd4 is playable, too. 14.xd4
c8 15.h3 d7 16.d1 e8 17.2f3 f8 18.b3
g6 (18...h6 as played in Calzetta-Kachiani, Crete
2007, is pointless, because, contrary to what both
ladies seem to have thought, it does not prevent
19.b2. Indeed, after 19...xe4 20.xe4 d5 the

18

Spanish Update

intermediate capture on h7 is not possible any

more, but 21.e5! leaves Black with problems
retrieving the sacrificed material in favourable
way.) 19.b2 g7 20.f4 e7 21.e1 c6=.
Blacks excellent development, the passive and
vulnerable placement of the c2-bishop as well
as the pressure against the e4-pawn should
compensate for the weakness of the d6-pawn.
14.h3
14.dxc5 xc5! 15.d4 c7 with active play for
Black.
14.d5 e8 is likely to transpose to 13.d5.
14.e1 does not cross Blacks intentions after
14...e8.
14...xf3 15.xf3 cxd4 16.xd4 e8 17.e1
d7 18.d3 f6=
This picture is familiar to us from the chapter
dedicated to the d3-system. Blacks better
development compensates for the small structural
defect.
19.e5?!
is not dangerous because of
19...f8!
Line C
9.h3 e8 10.00
10.g5 f8 (10...d5!? 11.exd5 xd4
12.cxd4 b4 13.d1 exd4 ...h6, ...xd5.)
11.a4 (11.e3 d5 12.exd5 xd5; 11.f3=)
11...d7=
10...f8 11.g5?!
11.g5 e7 12.a4 (12.d5 a5 13.c2 c6)
12...d7 13.d5 (13.axb5 axb5 14.xa8 xa8
15.d5 a5 16.a2 c6 17.a3 cxd5 18.exd5 e4
19.xb5 a6 20.c4 h6) 13...a5 14.c2 c6
11.d5?! a5 12.c2 c6
11.dxe5?! xe5! 12.xe5 xe5 13.f3 d5
Whites delay in development leaves him
struggling, Varavin-Kuzmin, Sudak 2002.
11...h6 12.xf6 xf6 13.d5 d7 14.dxe5
dxe5 15.a4 ad8!
The point behind this centralizing move will be
revealed later.
The mechanical (and unnecessary) defence of
the b5-pawn leaves White with some initiative.
15...ab8 16.axb5 axb5 17.a6 c5 (17...d4?!
does not work now because of 18.xf7! xf7

19.cxd4 exd4 20.xd4, for instance 20...c5

21.c2 b4 22.a5 and Black does not seem
to have sufficient compensation.) 18.a3 xa3
19.bxa3 Davies-Pavlovic, Vrnjacka Banja 1988.
16.axb5 axb5 17.a3
17.a6?! d4! 18.xf7 xf7 19.cxd4 exd4
Here is the difference! The d4-pawn is taboo now.
20.xd4? c8!
17.xb5 e7 18.e2 (White needs to prevent
...xh3 by defending the knight.) 18...xd5
19.exd5 e4 20.d4 g5
17...e7 18.b3 g6
Black has the pair of bishops and a very
harmonious placement of pieces.
Line D
11.bd2 a5 12.c2 c5 13.d1
13.e1 puts less pressure on Black and allows
him develop in the spirit of the Petrosian system
without moving the queen at all. 13...d7 14.f1
c4 15.g3 g6=
13.b4?! leaves Whites centre hanging after 13...
cxb4 14.cxb4 c6
13...e7
The most direct way to force White take a
decision in the centre.
13...c7 is also playable, but White can
sacrifice a pawn in order to maintain the tension.
14.f1!? exd4 (14...g6?! 15.g5! g7 16.dxe5
dxe5 17.xf6 xf6 18.e3 b7 19.a4)
15.cxd4 b7 16.g3 xe4 17.xe4 (17.xe4
d5 18.f6 gxf6 19.d3 e4 20.dxc5 xc5
21.f1) 17...xe4 18.xe4 d5 19.f6 gxf6
20.d3 c4 21.f5
13...d7 14.f1 c4 looks entirely playable
as well.
14.d5
14.dxc5 dxc5 15.f1 c4 16.g3 g6 17.b3
d6 18.e3 c4 19.c5 c7 20.b4 a5=; 14.a4
d7= transposes to 11.a4.
14...h5! 15.f1 g6 16.g5
16.a4 leaves the b3-square after 16c4.
16...c7=
Black intends to regroup with ...g7, ...d7,
eventually ...c4. At the right moment, the h5knight would jump to f4, even if this would mean
sacrificing a pawn for activating his g7-bishop.

Mihail Marin

Line E
11.a4 d7
11...b8?! 12.axb5 axb5 13.g5 d5 (Black
is forced to open the centre because 13...e7?
14.d5! leaves the knight trapped. We can see here
the drawbacks of leaving White with the control
of the a-file.) 14.xf7! In the style of Morphy!
14...xf7 15.exd5 exd4 16.dxc6 e6 17.xe6
xe6 18.f3 d5 19.xd5 xd5 20.cxd4 xc6
21.d2 c2 22.f3 Vitiugov-Pokazanjev,
Nojabrsk 2005.
11...exd4?! 12.axb5! xe4 13.xf7! VitiugovYamilov, St Petersburg 1999.
12.bd2
12.d5 e7 (12...a5 13.c2 c6) 13.c4 c6
(13...g6 14.axb5 axb5 15.xa8 xa8 16.cxb5
h5 17.c3 gf4 18.xf4?! White should
have refrained from this exchange, in order to
leave Black with the chronic problem of finding
a job for the superfluous knight. 18...xf4
19.e3 g6 20.h2 h6 21.g1 g7 22.f3
f5 Smirnov-Tarlev, Alushta 2004.) 14.g5 (14.
axb5 axb5 15.xa8 xa8 16.g5 bxc4 17.xc4
cxd5 18.exd5 exd5) 14...bxc4 15.xc4 cxd5
16.exd5 g6 17.c3 (17.xa6?! h6 18.xf6

19

xf6) 17...h6 18.xf6 xf6= 19.xa6? f4

20.c4 xh3
12...a5 13.c2 c5 14.d1
Here, too 14.b4 is premature because of 14...
cxb4 15.cxb4 c6
14...e7
This line is important for the move order
starting with 11.bd2.
However, in this concrete position, 14...c7
is an important alternative. 15.d5 (15.axb5 axb5
16.f1 b3 17.xa8 xc1 18.xc1 xa8;
15.f1 exd4 16.cxd4 bxa4) 15...h5 16.f1
g6=
15.f1
15.dxc5 dxc5 16.f1 c6 17.g3 g6=
15...exd4 16.cxd4 xe4
It appears that Black can embark this line
without fearing the pressure along the a-file.
17.axb5 axb5 18.xe4 xe4 19.xe4 xe4
20.g3
20.d2 b4!
20.dxc5 a4 21.xa4 bxa4 22.cxd6 c4=
20...ee8 21.dxc5 b3 22.xa8 xa8 23.e3
a1!=
The tension will be released completely and,
having the pair of bishops, Black is not worse.

20

Spanish Update

Update to Chapter 5

The Yates Variation 10.d5

Just a couple of weeks after the first edition of
the book was published, I had the occasion to
convince myself about the truth contained by the
final part of my conclusion.

Jakovenko Marin
Torrelavega 2007

Recently, I analyzed in depth for the Spanish

magazine Jaque a couple of Jakovenkos games
from the Poikovsky tournament and I must say
that I remained quite impressed. Among others,
his treatment of the Classical Najdorf (with
white) is very effective. Given the aforementioned
structural similarity between the Yates Spanish
and the Classical Najdorf, one could not dream
about a better suited opponent to test his Spanish
repertoire against.

1222222223
4t+v+ Tl+5
4+ + VoOo5
4o+wO M +5
4+o+ O + 5
4 +m+p+ +5
4+ P +n+p5
4pPb+ Pp+5
4R BqRnK 5
7888888889

In the diagrammed position, Jakovenko played

16.a4
Lately, I experience problems with my memory,
but during the game I remembered quite clearly

that my main line went 16.g3 e8 17.a4 e6

and hoped that his move would just lead to
transposition.
16...e6

1222222223
4t+ + Tl+5
4+ + VoOo5
4o+wOvM +5
4+o+ O + 5
4p+m+p+ +5
4+ P +n+p5
4 Pb+ Pp+5
4R BqRnK 5
7888888889

17.g5!? d7 18.b3 a5
Retreating with the knight to the more natural
b6-square with 18...b6 would actually leave it
slightly misplaced after 19.a5 c8.
19.d2

1222222223
4t+ + Tl+5
4+ +vVoOo5
4o+wO M +5
4Mo+ O N 5
4p+ +p+ +5
4+pP + +p5
4 +bB Pp+5
4R +qRnK 5
7888888889

Mihail Marin

I was convinced that I had never seen this

position before. Therefore, I was surprised to find
out that I had analyzed it in the book, although it
is given as a sideline.
19...fc8
Actually, only this is the new move and it seems
more logical than 19...c7, as given by me on
the basis of a game Nijboer-Adams, Wijk aan
Zee 1992. It is true that the queen retreat creates
the threat ...bxa4, practically forcing White to
release the tension with 20.axb5, but after 20...
axb5 21.e3 fc8 22.a2 Black had to adjust
the position of the queen with 22...b7, which
causes a loss of time. The conclusion of my
analysis was that after 23.d5 d8 the position
is balanced anyway.
20.e3
I was not familiar with the plan involving the
knight jump to e3, but I intuitively felt which
structure my opponent was aiming for. In the
Najdorf game from my aforementioned article
for Jaque, Jakovenko played a relatively early d5
and soon obtained an overwhelming advantage
on the queenside.
20...h6
But probably this move is not the best. I wanted
to return with my bishop to f5 in order to question
the viability of the knights placement on e3 in
view of the vulnerability of the e4-pawn.
It would have probably been better to refrain
from weakening the kingside and play 20...b7,
for instance 21.axb5 axb5 22.d5 d8 with an
extra-tempo compared to the book-variation.
21.f3
In case of 21.d5 xd5 22.exd5 Black can
avoid needless complications with 22...c7.
21...e6
Starting with this moment I gradually
understood that Whites plan has consistence
both strategically and tactically.
For the time being, this feeling had rather
vague contours. It is easy to see that 21...xe4?
does not work because of 22.axb5 axb5 23.d5!
xd5 24.xe4 followed by xa8 and b4 winning
lots of material.
22.h4!

21

1222222223
4t+t+ +l+5
4+ + VoO 5
4o+wOvM O5
4Mo+ O + 5
4p+ +p+ N5
4+pP N +p5
4 +bB Pp+5
4R +qR K 5
7888888889

But now the feeling became more distinct. I

expected that the other knight would occupy the
f5-square, but after 22.f5 f8 the f3-knight
would remain passive. The attempt to activate it
with 23.h2 (planning g4) would release the
pressure against the centre for just a moment,
allowing Black obtain strong counterplay with
23...b4 24.c4 xf5 25.exf5 d5!
Jakovenkos plan is more consistent: he intends
to occupy both critical squares (d5 and f5) with
his knights.
22...f8
Once again, he had worked out the tactical
nuances rather well. After a long thought I
decided that 22...xe4 was not advisable because
of 23.axb5! (The point behind this exchange
will become clear at the end of the variation.)
23...axb5 24.xe4 (24.ef5 is not dangerous
because of 24...xh4! for instance 25.xe4 d5
26.g4 f6 27.c2 e4, when the possibility of
capturing on h6 with check does not compensate
White for Blacks advantage of space.) 24...xe4
25.ef5 Here, I mainly checked the queen
sacrifice 25...xf5 (Objectively, 25...b7 is
better, but after 26.g4 Whites pressure is
unpleasant. Here is a sample line: 26...g5
27.xg5 hxg5 28.xg5 f6 29.g6 Strategically,
Black seems to have a good position, but he
cannot free his kingside easily. One of the main
problems is that ...f7? is impossible because
of h6. 29...d7 30.e3 xb3 31.d1 and
Blacks position will collapse soon.) 26.xe4 xe4
27.g4 and Blacks pieces are hanging.; I was
not too attracted by 22...g6 because of 23.hf5!
gxf5 (23...f8 transposes to a position from the

22

Spanish Update

next note.) 24.exf5 d7 25.g4 with dangerous

attack.
23.hf5 b6!?
The idea of this slightly mysterious move is
to consolidate the king side with ...d8 in case
of emergency. It took less time than on the
previous move to establish that 23...xe4? was
out of question. After 24.g4 xd2? (24...f6
is better, but after 25.xh6 h8 26.h4 White
has a promising attacking position without the
necessity of making major material investments.)
25.xh6 h8 26.xf7 g8 (26...xf7
27.h4 leads to mate, too.) 27.h5! and Black
is lost. The immediate threat is h7# and the
knight is taboo because of h7#. After 27...e4
28.g5 there is no defence either, partly because
the reserve knight is useful on e3 by preventing
28...f5.
By this time I started experiencing some
problems with the time. Therefore, I discarded
23...g6 on general ground: the kingside position is
weakened too much. I believe that my evaluation
was correct. Here is a sample line 24.d5!?
(Actually, the threat ...gxf5 is not real, which
means that White could make a strengthening
move still. For instance, 24.f3 and if
24...gxf5 then 25.xf5 or 25.g3 followed by
26.exf5.) 24...xd5 25.exd5 xd5 26.xh6
xh6 27.xh6 xg2 Both sides have managed
to considerably weaken the square placed right
in front of the enemy king, but whose attack
will be faster and more efficient? 28.g4! I
believe that it is easy to guess the answer already.
28...f3 29.h4 With the brutal threat f6.
29...e4 (29...xc3?, with the intention of meeting
30.f6 with 30...e4 is refuted by 30.e4! xe4
31.f6 with mate to follow.) 30.g5 f5 31.f6
h5 32.d1. White was a winning attack on
dark squares.
24.d5
A new surprise. I expected 24.f3, against
which I had designed 24...d8 anyway.
24...d8
I was about to play 24...xd5, when, just as
if prompted from above, I noticed the deadly
intermediate move 25.xh6!! (I would not
have mind the position arising after 25.exd5
xf5 26.xf5 e8, but certainly would have

refrained from 25...xd5 because of 26.xh6!

gxh6 27.g4 followed by f5.) After 25...gxh6
26.exd5 Blacks position is not defensible, for
instance 26...d7 (26...xd5 would transpose
above.) 27.h5. I still tried to find a defence
here with 27...d8 28.xh6 f6 but when I saw
29.h7! I dropped the whole thing.
25.f3

1222222223
4t+tW Vl+5
4+ + +oO 5
4o+ OvM O5
4Mo+nOn+ 5
4p+ +p+ +5
4+pP +q+p5
4 +bB Pp+5
4R + R K 5
7888888889

Another unexpected move. I was aware that

I had not guessed too many of my opponent's
moves after the opening and started feeling that
we were playing different games in fact.
While he was thinking, I mainly investigated the
consequences of 25.c4 and drew the conclusion
that 25...b7 would be OK for Black.
In fact, Jakovenkos move is very logical. By
creating the threat xh6, he forces Black release
the tension, when his favourite Najdorf structure
would suddenly arise on board. It is remarkable
that, even knowing his positional tastes, I failed
to predict concrete moves. Sometimes, it is not
easy to put up together all the information one
knows under the pressure of time.
25...xd5 26.exd5
Now, 26.xh6 gxh6 27.exd5 d7 28.h5
would be less dangerous because of 28...f5 and
the queen is close enough to prevent the disaster.
26...xf5 27.xf5
In case of 27.xf5 I would have played
27...xb3! instantly, without checking whether
Black loses an exchange or not. The strategic
gain on the queenside would have obviously
offered good compensation.
27...g6 28.g4

Mihail Marin

1222222223
4t+tW Vl+5
4+ + +o+ 5
4o+ O +oO5
4Mo+pO + 5
4p+ + +q+5
4+pP + +p5
4 +bB Pp+5
4R + R K 5
7888888889

Up to this moment, I considered that I

had managed to keep things under control in
satisfactory way. Black has a strong kingside
majority and should count on a long term
advantage on this wing, although for the
moment it is not easy to advance the pawns. I was
hesitating between such neutral moves as 28...h5
and 28...g7, when, suddenly, I understood my
opponents idea: axb5 followed by b4, when the
black queenside would be in big danger. It did
not take too much time to understand that such
an evolution would offer few hopes for successful
defence and I became very impressed about my
opponents anterior play.
Now that the over-the-board emotions have long
calmed down, I can express the psychological
situation more clearly. It is my firm belief that
9.d4 is slightly premature objectively speaking.
The same applies to 10.d5. This does not mean
that Black is better after any of these moves,
but he should be entitled to count on reaching
a viable position with more ease than in the
main lines based on 9.h3 followed by 10.d4 or
11.d4. At the same time, the strategic and tactical
complexity of the position in the range between
the 20th and 25th moves clearly reminded me of
Karpovs brilliant games in the classical variations
of the Ruy Lopez. Obviously, there was a point of
discontinuity in my whole thinking process.
When, later that same day, I complained about
this to my team mates, Jacob Aagaard gave me a
very wise consolation: The Ruy Lopez Opening
is enormously complicated, and so is chess.
Indeed, sometimes it is not easy to find the
absolute truth during the game. The comments

23

you are reading are fruit of additional analytical

work, but they only justify Whites strategy and
fail to spot Blacks mistake yet. Where did he
allow White undeserved chances to complicate
the fight from what should be relatively easy
to handle position? Even if some improvement
will be found in the range from the 19th to the
20th moves, Black should still be unsatisfied: this
is way too complex. I felt that I will either have
to find a relatively early deviation for Black or
convince myself of the truth that 9.d4 followed
by 10.d5 can be equally complex as 9.h3. After
all, it is never too late to learn something new
While preparing the updates for the second
edition, I finally found what seems to be an
adequate answer (see below). For the moment,
I will only add that, with less than 10 minutes
left to reach the 40th move, I decided to leave
the abstract questions for later and looked for a
concrete solution to my immediate problems.

1222222223
4t+tW Vl+5
4+ + +o+ 5
4o+ O +oO5
4Mo+pO + 5
4p+ + +q+5
4+pP + +p5
4 +bB Pp+5
4R + R K 5
7888888889

28...b4!?!
The best practical decision and the only way to
ensure the knight stability.
29.xb4 ab8 30.a3
This is the position I had aimed for when
sacrificing the pawn. Whites queen has been
forced to temporarily occupy a passive position
and the queenside majority is anything but easy
to advance. In the meanwhile, Black can freely
strengthen his position on the other wing.
Objectively speaking, his compensation should
not be 100% sufficient, but White is confronted
with unexpected practical problems. The position
remains equally complicated as before, but the

24

Spanish Update

typical ideas were probably not within the range

of Jakovenkos best knowledge, which eventually
led him to a (somewhat undeserved) defeat.
Let us return to the position after Whites 19th
move.

1222222223
4t+ + Tl+5
4+ +vVoOo5
4o+wO M +5
4Mo+ O N 5
4p+ +p+ +5
4+pP + +p5
4 +bB Pp+5
4R +qRnK 5
7888888889

Both 19...c7 (Adams) and 19...fc8 (Marin)

reveal Blacks desire (or impatience) to put pressure
on Whites position or to undertake concrete
action. The game against Jakovenko made me
look at the position with different eyes. Whites
plan to install his knights on d5 and f5 is most
disturbing and, if possible, should be prevented
by any means. This is how the prophylactic
move
19...fe8!
came to my mind. With the bishop well
defended, 20.e3 loses a pawn, possibly with
some compensation for White, but not more
than that.
20.g3 h6 21.f3 e6 transposes to the main
line.
After 20.c4 b7 21.cxb5 axb5 22.a5 Black
should cut the a5-pawn off his colleague with 22...
b4! for instance 23.xb4 h6 24.f3 d8 25.d2.
Not really defending the a5-pawn, because the
c2-bishop is hanging. 25...xa5! 26.xa5 xa5
27.xa5 xa5. The position is dynamically
balanced and could eventually simplify to a draw.
Here is a possible continuation: 28.e3 b6
29.b4 c6 30.xd6 xb4=
20...h6 21.f3
This is a first moment when the rook proves
useful on e8. The intermediate 21.d5 is not
dangerous at all.

1222222223
4t+ +t+l+5
4+ +vVoO 5
4o+wO M O5
4Mo+ O + 5
4p+ +p+ +5
4+pP Nn+p5
4 +bB Pp+5
4R +qR K 5
7888888889

21...xe4!
As we shall see, this move is entirely possible
now.
22.axb5
The tempting 22.xe4 xe4 23.c4 fails to
23...g6! 24.xa5 xh3 25.g3 g4 followed
by ...e4. 22.c4 xd2 23.xd2 offers White nice
control on light squares, but Blacks compact
structure ensures him against troubles after, say
23...b6 24.d5 d8.
The same applies after 22.b4 xd2 23.xd2
c4 24.dxc4 bxc4 25.d5 f5.
22...axb5 23.d5
We can see the difference now: Black has time
for the intermediate
23...xd2 when after 24.xe7 xe7 25.xd2
a6 26.a3 b7
Blacks temporary lack of coordination should
enable White to maintain equality, but not more
than that. For instance,
27.e4 b6 28.xa6 xa6 29.xb7 xb7
30.xd6 e6 followed by ...c6, ...f6 or the
more daring ...g6, according to Whites answer.
I believe that we have enough elements to stick
to the conclusion from the previous edition

Mihail Marin

25

Theory
Line A
19...fc8 20.e3 h6
20...b7!? 21.axb5 axb5 22.d5 d8 with an
extra-tempo compared to the line 19...c7.
21.f3
21.d5 xd5 22.exd5 c7!?=
21...e6
21...xe4? 22.axb5 axb5 23.d5! xd5
24.xe4 followed by xa8 and b4 winning lots
of material.
22.h4!
22.f5 f8 23.h2 ( g4) 23...b4 24.c4
xf5 25.exf5 d5!
22...f8
22...xe4 23.axb5! axb5 24.xe4 (24.ef5
xh4! 25.xe4 d5 26.g4 f6 27.c2 e4)
24...xe4 25.ef5 xf5?! (25...b7 26.g4
g5 27.xg5 hxg5 28.xg5 f6 29.g6 One of
the main problems is that ...f7? is impossible
because of h6. 29...d7 30.e3 xb3
31.d1-) 26.xe4 xe4 27.g422...g6 23.hf5! gxf5 (23...f8 transposes
to a position from the next note.) 24.exf5 d7
25.g4
23.hf5 b6!?
...d8.
23...xe4? 24.g4 xd2? (24...f6 25.xh6
h8 26.h4) 25.xh6 h8 26.xf7
g8 (26...xf7 27.h4 leads to mate, too.)
27.h5!23...g6 24.d5!? (24.f3 gxf5 25.xf5
or 25.g3 followed by 26.exf5.) 24...xd5
25.exd5 xd5 26.xh6 xh6 27.xh6 xg2
28.g4! f3 29.h4 f6. 29...e4 (29...xc3?,
with the intention of meeting 30.f6 with 30...
e4 is refuted by 30.e4! xe4 31.f6 with mate
to follow.) 30.g5 f5 31.f6 h5 32.d1-

24.d5
24.f3 d8
24...d8
24...xd5? 25.xh6!! (25.exd5 xf5 26.xf5
e8, but not 25...xd5 26.xh6! gxh6 27.g4
followed by f5.) 25...gxh6 26.exd5 d7
(26...xd5 would transpose above.) 27.h5
d8 28.xh6 f6 29.h7!
25.f3
25.c4 b7
25...xd5 26.exd5
26.xh6 gxh6 27.exd5 d7 28.h5 f5.
26...xf5 27.xf5
27.xf5 xb3!
27...g6 28.g4
Jakovenko-Marin, Spanish team championship
2007.

Line B
19...fe8! 20.e3
20.g3 h6 21.f3 e6 transposes to the main
line (16. g3 e8 17.a4 e6).
20.c4 b7 21.cxb5 axb5 22.a5 b4! 23.xb4
h6 24.f3 d8 25.d2 xa5! 26.xa5 xa5
27.xa5 xa5 28.e3 b6 29.b4 c6 30.xd6
xb4=
20...h6 21.f3 xe4! 22.axb5
22.xe4 xe4 23.c4 g6! 24.xa5 xh3
25.g3 g4 followed by ...e4.
22.c4 xd2 23.xd2 b6 24.d5 d8
22.b4 xd2 23.xd2 c4 24.dxc4 bxc4
25.d5 f5
22...axb5 23.d5 xd2 24.xe7 xe7
25.xd2 a6 26.a3 b7 27.e4 b6
28.xa6 xa6 29.xb7 xb7 30.xd6 e6
followed by ...c6, ...f6 or the more daring
...g6, according to Whites answer.